The Ultimate Averaged Chart - The BBC Chart Re-Imagined

[MrTibbs]

Here is the next Ultimate Averaged Chart for Week Ending October 6th 1956

Here are all '' the uppers, the downers, the just hanging 'arounders '

The Ultimate Averaged Chart - Week Ending October 6th 1956							NME	MM	RM	Total
Last	This		The Sound Survey			Stores	65	20	60	Points
Week	Week		The Top 30 Singles Chart					TOP	30	Scored
1	1	Lay Down Your Arms - Anne Shelton					1	1	1	4350
6	2	A Woman In Love - Frankie Laine					3	2	2	4140
2	3	Whatever Will Be Will be - Doris Day					2	4	3	4105
5	4	Hound Dog - Elvis Presley					4	3	4	3935
20	5	Giddy-Up-A-Ding-Dong - Freddie Bell and The Bell Boys					6	5	5	3705
3	6	Rocking Through The Rye - Bill Haley and His Comets					5	6	6	3690
4	7	Ying Tong Song / Bloodnok's Rock'n'Roll Call - The Goons					7	8	7	3460
10	8	Only You - The Platters (B)						7	9	3295
7	9	The Great Pretender - The Platters (A)						13	8	3073
8	10	Bring A Little Water Sylvie - Lonnie Donegan (B)						10	11	2953
14	11	Rock Around The Clock - Bill Haley and The Comets					12	16	10	2795
17	12	Born To Be With You - The Chordettes					11	11	16	2600
12	13	The Saints Rock'n'Roll - Bill Haley and His Comets					16	9	12	2555
13	14	Dead Or Alive - Lonnie Donegan (A)						15	13	2505
11	15	Walk Hand In Hand - Tony Martin					10		17	2205
21	16	Razzle Dazzle - Bill Haley and The Comets					13	20	18	2170
19	17	When Mexico Gave Up The Rumba - Mitchell Torok					19	11	15	2140
16	18	Why Do Fools Fall In Love - The Teenagers					17	20	19	1850
24	19	See You Later Alligator - Bill Haley and His Comets					20		14	1735
9	20	A Sweet Old Fashioned Girl - Teresa Brewer					15	18		1300
15	21	Mountain Greenery - Mel Torme					13			1170
29	22	More - Perry Como (A)					18	17		1125
22	23	I'm In Love Again - Fats Domino					24	13		815
NEW	24	A Woman In Love - Ronnie Hilton							20	660
RE	25	You Are My First Love - Ruby Murray					21			650
29	26	More - Jimmy Young					22			585
28	27	Be-Bop-A-Lula - Gene Vincent					23			520
18	28	Serenade - Slim Whitman					27	19		500
RE	29	Love Me As Though There Were No Tomorrow - Nat King Cole					25			390
RE	30	Walk Hand In Hand - Jimmy Parkinson					26			325
		Ten Thousand Miles - Michael Holliday					28			195
26		Glendora - Perry Como (B)					29			130
		Glendora - Glen Mason					30			65
		Only You / The Great Pretender - The Platters					8
		Bring A Little Water Sylvie / Dead Or Alive - Lonnie Donegan					9
23		Heartbreak Hotel - Elvis Presley
24		By The Fountains Of Rome - Edmund Hockridge
27		I Want You I Need You I Love You - Elvis Presley

[RokinRobinOfLocksley]

Here's another thought I haven't done any research on: let's say you have 3 charts. Some records will appear on all 3, some only on 2, some only on 1. Maybe all the 3 chart records should appear on the combo chart first at the top, followed by all the 2 chart records in the middle, followed by all the 1 chart records at the bottom. Using weighted points to sort them within each category.

Just an alternative eye deer...

[RokinRobinOfLocksley]

I was doing inverse points like this: For the Top 50 chart, chart position #1 = 50 inverse points, down to chart position #50 = 1 inverse point. Or, inverse points = 51 - chart position.

For the two Top 30 charts, the same thing for chart positions #1 down to #30: #1 = 50 inverse points, down to #30 = 21 inverse points. Then awarding inverse points to the missing records at what would be positions #31 down to #50:

a = 0 inverse points for all of them
b = 20 inverse points for all of them
c = 10.5 inverse points for all of them
d = 51 - the average chart position for the charts the missing records appear on

Something like that. It was years ago. I just pulled up my old spreadsheet, and it's got so many scenarios on it (somewhere around 50 or so), I'd be super straining my brain if I were to try and figure it out again, ha.

[Splodj]

Robin - I assume in (a) you are counting up as in UAC and in (b) you are counting down as in BBC. If I have misunderstood please set me right!

[Splodj]

I would have thought (d) is the best. I don't think it is "assuming data that does not exist", it is maximising the data that does exist. When MM was compiling a Top 50 I reckon this was very valuable data. (Not so sure about RR - particularly in the early low sample days!! - but weighting could take care of that,)

(a) and (b) produce exactly the same results as each other when I test them.

I think the description in (a) of "not awarding any extra points" is misleading. What matters is that you are awarding a fixed number of points (zero points) and that is the same as awarding a fixed number of points (31 points) counting the other way.

Zero is not a neutral concept (in this context) it is a number with +1 above it and -1 below it. You can pivot the zero anywhere. Pivoting it at number 31 does not make it any more neutral than pivoting it at (say) 17. If you pivot it at 31 that means number 32 is -1 (minus one) point, number 33 is -2 points and so on. So if you found out that the real position was number 33, under the BBC system you should award 33 points and under the UAC system you should award -2 points. By awarding the default of 31 points (BBC) and 0 points (UAC) you are out by exactly the same number of points under both systems.

[RokinRobinOfLocksley]

For what it's worth, I did some limited research on this awhile back, something similar to combining a Top 30 chart with 2 Top 20s. Actually, I think I looked at combining a Top 50 with 2 Top 30s. Same deal.

I tried 4 methods for the missing records #31 to #50:
a--not awarding any extra points for the missing records
b--using #31 for all the missing records
c--using the average chart position #40.5 for all the missing records
d--calculating an average of the other charts for the missing records

Then I did the combos. I think I did a simple combo with no weighting of record shops, and then a combo with weighting of record shops.

The end result of this limited research (if my memory is correct) is that all 4 methods were relatively close to each other, that only a few records were affected at the lower end of the combo charts. But for methods b, c, and d, you are assuming things that are not in evidence.

b is the worst of the lot because you are adding more total points to the chart, you can't have 20 records tied at #31. A 50 position chart would normally have 1275 total inverse points, before applying record shop weightings. But a 30 position chart, with 20 records ranked at #31 would give a total of 1465 inverse points, for an extra 190 points. This throws things off, giving an advantage to those records that are missing from positions #31 to #50 on those charts.

c and d are less bad, but still you are assuming data that does not exist.

Philosophically speaking, 'a' is the most logical mathematical solution, you are going with only the data you have. b is not valid, and there are only minimal combo differences compared to c and d. Of course each week's charts will give different results, but overall 'a' is the way to go.

[Splodj]

To be fair, for most of the 'composite' period the BBC calculated 30 but broadcast only a Top 20. When we look at their Top 30s in that file it is easy to forget that the 21-30 section, which contained the bulk of such 'distortions', was only for their private use.

[MrTibbs]

If you have NME positions 21-30 why not use them? Naturally it is imperfect because NME is only shining some light on 21-30, but surely some light is better than none!



This is an interesting possibility. If all three charts were Top 20s it would be unfair to assume that an absentee is at number 21 - because that is the highest possible position it could be. You could do some research and determine an average position. If this were (say) 25 you could then treat an absentee as at number 25. So you would then award each absentee 25 points (BBC method) or minus 4 points (UAC) method, instead of 21 and zero respectively.

But I don't see how applying the same score to positions 21-30 (whether 21 or 25) can be superior to applying the actual NME positions.

Or am I missing something? (again!)

I totally agree with you there Splodj because that is my reasoning as well. At least NME's 21 to 30 positions were factual so at least the bottom ten of the UAC is based on fact not supposition. To my mind that's the error the BBC made, assuming every record not in an individual chart be awarded points as if it were #21 or #31 which all of them couldn't be. It totally distorted, especially the lower positions, in the BBC chart.

[Splodj]

If you have NME positions 21-30 why not use them? Naturally it is imperfect because NME is only shining some light on 21-30, but surely some light is better than none!

or you could choose an average points figure

This is an interesting possibility. If all three charts were Top 20s it would be unfair to assume that an absentee is at number 21 - because that is the highest possible position it could be. You could do some research and determine an average position. If this were (say) 25 you could then treat an absentee as at number 25. So you would then award each absentee 25 points (BBC method) or minus 4 points (UAC) method, instead of 21 and zero respectively.

But I don't see how applying the same score to positions 21-30 (whether 21 or 25) can be superior to applying the actual NME positions.

Or am I missing something? (again!)

[membranemusic]

Mr Tibbs, yes I totally accept your point of view. The issue of lower figures 'invention' was a real problem when I was busy in the late 1960s doing the straight average Top 50 every week.

With that in mind, I nonetheless had the nerve to juggle the figures for the UAC chart of 29 Sep 1956.. For those who are interested, the Platters double come in at No. 6 on jiggled joint points, and Lonnie Donegan up to No. 7. This was on my Method 2, but the same result happens with both on Method 1. Further down, Perry Como does the jiggle at No. 24 including RM compensation. But we have only 29 records ion the Top 30!

I won't be doing this every week, promise.

[Splodj]

Splodj, the most egregious example of the BBC using Method 1 for double –sided hits was “The Next Time/Bachelor Boy”, which on 29^th December 1962 and January 26th 1963, should have been No. 1 on the BBC chart, though fortunately it did not affect intervening weeks.

Without doing the calcs:
“Rock a Hula Baby/Can’t Help Falling In Love”: 3^rd March 1962, 10th March 1962, 17^th March 1962 - possibly more weeks at No.1 on the BBC.

“Blue Bayou / Mean Woman Blues” - Roy Orbison: 12 Oct 1963, 19^th Oct 1963, 2^nd Nov 1963 , 9^th November 1963 - possibly higher than No. 4 on the BBC, although there was a strong No. 3?

However I think the UAC reflects a fair outcome in these cases.

Yes, you are right and I am wrong. I even criticised the BBC on this thread for its 'split' methodology and said it should have used the same system it did for EPs. My memory then tricked me into thinking it had!

Of course by the time the BBC chart came along NME was the only splitter. We cannot tell how a BBC chart would have treated the issue at this time when most were splitting. I don't think the 'highest placing' method would have been workable and suspect they too would have split.

[MrTibbs]

Hi membrane. Your thoughts on points below #20 is valid, the only downside being we don't have a clue what positions any record actually held below 20 on RM or MM. Originally I thought about using the same system the BBC used ie considering all records below #20 on RM and MM to be at #21 to equalise points but then took the view as it wasn't factual and as all records couldn't hold that position it would actually make the bottom ten of the UAC more a guesstimate than it already is so just went with the straight positions. As we truly have no evidence that Mitchell Torok would definitely have been in an expanded RM Top 30 it would not be fair to assume so and credit it with points (it wasn't in NME's Top 30 either) .

Because of this I've always said that between 1955 and late 1962 when MM and Disc also moved to a Top 30 as well to take the Top 20 as factual and consider the bottom ten with a tongue in cheek approach.

That said I sincerely appreciate you raising this valid point and hope my explanation reassures you.

[membranemusic]

Sorry to raise another query on the same chart 22 Sep 1956 - not a double-sided issue. This is almost trivial, but ....

Should Mitchell Torok not be No. 28? As I see it, his record entry in MM is being under-compensated by the lack of Nos 21-30 on the RM chart. This under-compensation affects all records equally apart from those that get a placing in the other Top 20. If you give UAC nos 21-29 [all NME entries] another 600 points as though they were all joint No. 21 on RM [or you could choose an average points figure], plus Torok also, then Torok’s score is 860, enabling him to pip D Hughes in a tie break by virtue of being much higher on MM (No. 18).
Just an opinion!

[membranemusic]

Splodj, the most egregious example of the BBC using Method 1 for double –sided hits was “The Next Time/Bachelor Boy”, which on 29^th December 1962 and January 26th 1963, should have been No. 1 on the BBC chart, though fortunately it did not affect intervening weeks.

Without doing the calcs:
“Rock a Hula Baby/Can’t Help Falling In Love”: 3^rd March 1962, 10th March 1962, 17^th March 1962 - possibly more weeks at No.1 on the BBC.

“Blue Bayou / Mean Woman Blues” - Roy Orbison: 12 Oct 1963, 19^th Oct 1963, 2^nd Nov 1963 , 9^th November 1963 - possibly higher than No. 4 on the BBC, although there was a strong No. 3?

However I think the UAC reflects a fair outcome in these cases.

[Graham76man]

Radio Luxembourg might also play a record by an artist from the USA that wasn't released in this country yet. So if you wanted the record by that artist you wouldn't have got it. Though you might have got the record if you lived in a port town, such as Liverpool, as they did have more imports coming in. The sheet music charts were sometimes a bit quicker than record industry at getting a song out. They certainly were in the late 40's and early 50's. Though the Sheet Music Charts tended to hold on to records, even if sometimes the distributors of records had stopped sending them out. I had a look at the sheet music chart for the 29 September and found two songs missing in the top 20. Who Are We? And Autumn Concerto.
The "More" track being new on the above chart had been in three weeks on the Sheet chart and was number 7. Showing a much slower release on record. The top two were flipped on the Sheet Chart.

[MrTibbs]

I think another important factor at this time is that the record charts were still in the shadow of the 'Sheet Music Chart'. It was already well established and was given just as prominent a display in both NME and MM chart pages alongside their record charts. Also at this time Radio Luxembourg's Top Twenty Show had a massive radio audience and this remember was a broadcast of the 'Sheet Music Chart'.

So, the record buying public I believe were still asking for specific songs they heard on the radio as opposed to a record, so they would be asking for The Great Pretender or Only You when they went into their record shop. Some stores would therefore report on both sides independently when they gave their report to the relevant music paper while others would just add both together in their return.
The dealer returns had to be reflecting this confusion as the ones I have from '54 and '55 for RM showed some sides split and some combined on individual returns. Compilation would have been as much a nightmare for the music papers as it is for me just coping with three sets of figures.

So NME were possibly predominantly getting the 'add together' in their returns and ignored the split returns, and MM and RM were getting split returns and ignored joint returns. Who would know.

[Splodj]

split sides taking up two chart positions and denying other records an opportunity to chart.

And this why I think that when some charts split sides there is no 'perfect' way of doing a composite combined chart. In the RM chart above, even if you moved up the positions below 10 two places you still wouldn't know what numbers 19 and 20 should have been.

Has anyone done an analysis on which records were most affected?

NME denied Elvis (Rock-a-hoola/Can't help falling in love) number 1 and allowed Cliff (Next Time/Bachelor Boy) there for only 1 week. Perhaps most damaging was NME denying the Everleys (Bird Song) the top spot meaning it is missing from the 'official' pantheon of number ones.

[MrTibbs]

I actually find it strange that charts at that time actually did separate sides. When the NME launched their 'record chart' in 1952 you would have thought that is exactly what it would be but from the start the NME chart listed split sides which defied the principle of a 'record chart' as it really was still a song chart in reality.
Both RM and MM further perpetrated this practise on their arrival on the chart scene and as a result the 'record' charts of that era were a mess littered with split sides taking up two chart positions and denying other records an opportunity to chart.

Why oh why couldn't they all have stuck to a record is a record comprising of two sides, end of story.

[Splodj]

If you take the RM chart, Donegan is at 8 & 9 and The Platters are at 6 & 10 so, on a 'combined' basis, every position below 10 is two places lower than it should be.

So if you were compiling a composite all-combined chart would you jack up all those records two places?

[Splodj]

the BBC just took the easy way out for split sides and considered only the highest placed

I stand corrected. If they had done that at this time, when all charts were liable to separate, it would have produced bizarre results. With 'The Platters' above, the separators can't even agree on which side is the highest placed!

[MrTibbs]

Here is the next Ultimate Averaged Chart for Week Ending September 29th 1956

Here are all '' the uppers, the downers, the just hanging 'arounders '

The Ultimate Averaged Chart - Week Ending September 29th 1956							NME	MM	RM	Total
Last	This		The Sound Survey			Stores	65	20	60	Points
Week	Week		The Top 30 Singles Chart					TOP	30	Scored
1	1	Lay Down Your Arms - Anne Shelton					1	1	1	4350
2	2	Whatever Will Be Will Be - Doris Day					2	2	2	4205
3	3	Rocking Through The Rye - Bill Haley and His Comets					4	3	3	3995
4	4	Ying Tong Song / Bloodnok's Rock'n'Roll Call - The Goons					3	6	4	3940
16	5	Hound Dog - Elvis Presley					5	5	5	3770
11	6	A Woman In Love - Frankie Laine					8	4	7	3475
8	7	The Great Pretender - The Platters (A)						13	6	3258
9	8	Bring A Little Water Sylvie - Lonnie Donegan (B)						12	8	3125
7	9	A Sweet Old Fashioned Girl - Teresa Brewer					10	6	11	3065
10	10	Only You - The Platters (B)						11	10	2993
6	11	Walk Hand In Hand - Tony Martin					9	10	14	2870
12	12	The Saints Rock'n'Roll - Bill Haley and His Comets					11	16	12	2740
15	13	Dead Or Alive - Lonnie Donegan (A)						19	9	2665
19	14	Rock Around The Clock - Bill Haley and His Comets					13	14	13	2590
13	15	Mountain Greenery - Mel Torme					12	15	15	2515
5	16	Why Do Fools Fall In Love - The Teenagers					14	8	17	2405
14	17	Born To be With You - The Chordettes					15	8	16	2400
17	18	Serenade - Slim Whitman					18		20	1505
30	19	When Mexico Gave Up The Rumba - Mitchell Torok					23		18	1300
NEW	20	Giddy-Up-A-Ding-Dong - Freddie Bell and The Bell Boys					16	20		1195
NEW	21	Razzle Dazzle - Bill Haley and His Comets					17			910
24	22	I'm In Love Again - Fats Domino					29		19	850
26	23	Heartbreak Hotel - Elvis Presley					22	18		845
23	24	See You later Alligator - Bill Haley and His Comets					19			780
22	24	By The Fountains Of Rome - Edmund Hockridge					19			780
NEW	26	Glendora - Perry Como (B)					25	17		670
21	27	I Want You I Need You I Love You - Elvis Presley					21			650
RE	28	Be-Bop-A-Lula - Gene Vincent					24			455
NEW	29	More - Perry Como (A)					26			325
NEW	29	More - Jimmy Young					26			325
		Glendora - Glen Mason					28			195
18		I Almost Lost My Mind - Pat Boone					30			65
		Only You - The Great Pretender - The Platters					6
		Bring A Little Water Sylvie / Dead Or Alive - Lonnie Donegan					7
20		I'll Be Home - Pat Boone
25		The Wayward Wind - Tex Ritter
27		You Are My First Love - Ruby Murray
28		Love Me As Though There Were No Tomorrow - Nat King Cole
29		By The Fountains Of Rome - David Hughes

[MrTibbs]

Above three comments are interesting and thanks for those perspectives.

[QUOTE=membranemusic; I notice for the Goons you have factored in the difference between the MM top position (13), and the other best top position (NME & RM).
[/QUOTE]

That's not how I achieved that position for The Goons membrane. I simply ignored MM because it had split the sides and simply took an average of NME and RM and awarded that to MM. Also using that formula for The Platters like you explained placing it at #5 would just place it to equal the combined chart position on NME which I don't think would be realistic as there is no way of knowing what the combined positions would have been on MM and RM. By averaging both charts with split positions at least gives a middle point which at least is evidenced.

As setg 1 explains above I also suspect without any evidence that some sides were combined in NME but only the higher placed side mentioned. Let's be honest here, these fifties charts taken at face value are almost certainly less accurate than later years and those were only really a guide. BUT by combining them and taking an average for the UAC at least provides a more robust outcome. Combined has to be better than any individual chart.

As I said a couple of times before the fifties is by a country mile the most difficult era to work with because of the very difficult task involved in applying a formula to an ever changing landscape that is the various music papers charts sometimes combining sometimes splitting so there is no consistent structure to work with. Whatever system I use some record will be disadvantaged.
That said the approach I take I do think is the fairest. At this time the record chart was still very 'song' orientated as opposed to seeing it as a two sided record so given that most charts for the time being split the record this has to be taken into account. This is what the charts reflected at this time.
Ironically, MM and RM, then Disc, will also combine sides like NME , then come the sixties it will be NME who splits the sides.

If I remember correctly the BBC just took the easy way out for split sides and considered only the highest placed ignoring the lower which was unfair as it did not represent either side accurately. For EP's in the sixties when RR did not include them the BBC did much as I do at present and took an average from the other papers' charts and awarded this to RR to establish a chart position.

[Splodj]

I think until the situation settles down, and NME is the only separator, there are going to be difficulties either way. It is a credit to RM and MM that eventually they saw the light and combined.

As alluded to in the post above, one of the problems of separation is that unless both sides appear in the chart you never know how that record has been treated unless the chart compiler has a fixed policy of combining - which none of them did at this time.

Method 1 was the BBC method of just taking the 1^st side in each case, and ignoring the 2^nd-side chart position, producing some under-performing hits sometimes.

Do you have an example? In all the cases I have seen, the BBC ignored NME's separated positions and applied the other charts joint position.

[setg1]

I also suggested combining the sides, but most people seemed to disagree. Another problem is that sometimes a chart might combine both sides but still only list one of the songs. Record Mirror appears to have done this with the Goons single. I'm wondering if NME did the same with the Johnston Brothers single Hernando's Hideaway / Hey There in 1955, one of the records that started the discussion on what to do with the separately listed sides. The weekly positions for the songs on each chart:

NME	RM (HH)	RM (HT)
10	7	18
4	5	16
2	5	11
1	4	-
1	3	-
3	4	14
5	8	14
10	15	13
8	-	19
11	-	-
15	-	-

"Hey There" never appeared on the NME chart despite being close to the top ten in RM, while NME almost always gave "Hernando's Hideaway" a higher position than RM. It's impossible to say for sure, but I'd assume NME combined the points for both sides and then only credited them to one of the sides.

[membranemusic]

Understood – there’s a logic there based on the published perception of the 3 charts.
However, and this is just my opinion, I personally take a more record-business point of view in that 1 disc should have 1 chart entry even though the B-side is responsible for reported sales.
Normally this is not a problem, but is this era of September 1956 there is an interesting dichotomy for the chart compilers here. So how does one combine sales points? You have already achieved this with The Goons.
Method 1 was the BBC method of just taking the 1^st side in each case, and ignoring the 2^nd-side chart position, producing some under-performing hits sometimes.
Method 2 is factoring in points compensation for the 2^nd side. One has to be careful here, otherwise the record would be No.1. I notice for the Goons you have factored in the difference between the MM top position (13), and the other best top position (NME & RM). One could do the same with the Platters and Lonnie Donegan.

For 22 Sep 1956, the Platters double (with 4 + 1 places compensation) would achieve 3895 points and the No. 5 position, not Nos. 8 and 10. Lonnie Donegan (with just 1 + 1 places compensation) achieves 3335 points and would be No.9, not Nos 9 and 15. I hope my calcs are good. This also means that further down, records are moved up 2 positions. Is this fairer?
Just an opinion!